Naming Conventions

Each subroutine within KAPLIBS has a name of the form <prefix>_<name> where <prefix> is a prefix indicating which KAPPA library the routine belongs to, and <name> is a unique name for the routine. The prefix associated with each KAPPA library indicates something of the purpose of that library. The following prefixes are currently included within KAPLIBS:

AIF_

These are old routines which were used to access parameter values or temporary work space. They are gradually being replaced within KAPPA by more modern routines provided by the PAR and PSX libraries.

FTS1_

These are routines used to access FITS files and headers - KAPPA's own fitsio.

IRA_

These routines were initially part of the IRAS90 Astrometry library (hence the IRA acronym). They are used to gain access to WCS information stored within NDFs in the form of IRAS90 Astrometry extension. This form of WCS is now deprecated in favour of the NDF WCS component, accessed through the AST library.

KPG1_

These are other general-purpose routines which do not fall into such obvious groups. Even within this library, though, there are loose associations of routines, usually indicated by some common element within the routine name. The following are some of the more significant associations:

• Routines associated with use of the PGPLOT graphics package usually have names which start ``KPG1_PG''.
• Routines associated with use of the AGI graphics database usually have names which start ``KPG1_GD''.
• Routines associated with accessing or using WCS information usually have names which start ``KPG1_AS''.

Within all these groups, some routines have different versions for processing data with different numerical types. The names of such routines are identical except for the trailing one or two characters which indicate the data type processed. These one- or two-character codes are:

d

- Double-precision floating point

r

- Single-precision floating point

c

- Character

i

- Single-precision integer

w

- Word (usually two-byte integers)

uw

- Unsigned word (usually two-byte unsigned integers)

bw

- Byte (usually one-byte integers)

ub

- Unsigned byte (usually one-byte unsigned integers)

l

- Logical

Sometimes, routine names are documented as ending with a lower-case ``x''. This indicates that routines exist for various of the above numerical types. The actual routines names will not include the trailing ``x'', but will have one of the above codes in place of the ``x''.